On the mechanism of the palladium bis(NHC) complex catalyzed CH functionalization of propane: Experiment and DFT calculations

Article abstract of DOI:10.1002/chem.201403910

We report a detailed mechanistic study on the CH functionalization of alkanes by palladium complexes with chelating bis(N -heterocyclic carbene) (NHC) complexes. The experimental results are complemented by detailed DFT calculations, which allow us to rationalize the regioselectivity and the catalytic activity. The study includes a library of catalysts with different electronic and steric properties, kinetic data, and isotope effects. The combined experimental and computational results favor a mechanism involving organometallic palladium(IV) intermediates. Furthermore, it is shown that at high halide loadings a different mechanism is operative.

Full text of DOI:10.1002/chem.201403910

DOI: 10.1002/chem.201403910
Full Paper
&
C–H Activation
On the Mechanism of the Palladium Bis(NHC) Complex Catalyzed
CH Functionalization of Propane: Experiment and DFT
Calculations**
Dominik Munz and Thomas Strassner*[a]
Abstract: We report a detailed mechanistic study on the CH
functionalization of alkanes by palladium complexes with
chelating bis(N-heterocyclic carbene) (NHC) complexes. The
experimental results are complemented by detailed DFT cal-
culations, which allow us to rationalize the regioselectivity
and the catalytic activity. The study includes a library of cata-
lysts with different electronic and steric properties, kinetic
data, and isotope effects. The combined experimental and
computational results favor a mechanism involving organo-
metallic palladium(IV) intermediates. Furthermore, it is
shown that at high halide loadings a different mechanism is
operative.
acid (HOTFA) suffers from catalyst deactivation phenomena.[6]
Radical pathways, which are compatible with aqueous reaction
environments, usually suffer from low chemoselectivity and/or
the usage of unfavorable oxidants.[7]
Introduction
The functionalization of hydrocarbons has been one of the
most exciting areas of catalysis for decades. Likewise, the re-
finement of natural gas is becoming increasingly important be-
cause of the enormous shale gas boom in North America.[1] It
has been predicted that, unlike coal and crude oil, the share of
natural gas in global energy consumption is going to increase
considerably during the next 25 years.[2]
Palladium compounds were also studied in the context of
alkane functionalization[8] and were even considered for indus-
trial application.[9] We have shown that palladium catalysts,
which are stabilized by chelating bis-N-heterocyclic carbene
(NHC) ligands, exhibit very high catalytic activity with methane
in trifluoroacetic acid at 908C, using K2S2O8 as oxidant.[10] Later,
we evaluated the reactivity of a series of palladium bis(NHC)
catalysts with propane.[11]
Unfortunately, the gaseous state of natural gas renders its
transport challenging and billions of cubic meters of natural
gas are flared or vented annually. The CO-involving Fischer–
Tropsch process, which is commercially used by Sasol and
Shell, is very energy demanding as well. Therefore, the devel-
opment of an efficient chemical process, which ideally converts
methane directly to methanol, is highly desirable. Direct and
selective oxidative strategies do not suffer from those draw-
backs. Homogeneous transition-metal-mediated CH functional-
ization protocols appeared to be promising and have been
studied in detail by numerous research groups.[3]
We also modeled the mechanism of the palladium-bis(NHC)
complex catalyzed CH functionalization of methane by means
of DFT calculations.[12] We proposed, based on previous work
and the synthesis and reactivity of potential intermediates
(Figure 1),[13] that the reaction proceeds with a palladium cata-
lyst in the oxidation states of +II and +IV.
However, even 45 years after the seminal work of Shilov,[4]
the development of an efficient methane CH-functionalization
process using a reasonable reaction medium and cheap oxi-
dant remains elusive. The most prominent example constitutes
the platinum-based Catalytica system, which uses oleum as
solvent and oxidant.[5] Similarly, the aerobic functionalization of
methane with manganese and cobalt salts in trifluoroacetic
Figure 1. Synthesized potential intermediates.
The CH activation step probably takes place with a palladiu-
m(II) complex (Scheme 1) and is followed by a bromine-medi-
ated oxidation to produce a palladium(IV) alkyl species. The tri-
fluoroacetate ester is then released by reductive elimination
through external nucleophilic attack by a trifluoroacetate mole-
cule. The intermediately active oxidant bromine can be regen-
erated by oxidants like K2S2O8, which thereby produce the cat-
alytically potent complex 1 by an oxidative exchange of the
bromido for trifluoroacetato ligands.
[a] Dr. D. Munz, Prof. T. Strassner
Physikalische Organische Chemie
Technische Universitꢀt Dresden
01069 Dresden (Germany)
Fax: (+49)351 463-39679
E-mail: thomas.strassner@chemie.tu-dresden.de
[**] NHC = N-heterocyclic carbene.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201403910.
Chem. Eur. J. 2014, 20, 14872 – 14879
14872
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
n
ing palladium s complexes 2 (15.7 kcalmolꢀ1) or iso2 (15.3 kcal
molꢀ1). The subsequent CH bond activation (ts[n2>n3] and
ts[iso2>iso3]) is calculated to proceed by a cationic, intramolec-
ular transition state, preferentially at one of the terminal
¼
methyl groups of propane (DDG =5.0 kcalmolꢀ1). Opposed
to the kinetic selectivity, the iso-product iso3 is thermodynami-
n
cally more stable than the n-product 3 by 0.4 kcalmolꢀ1. The
n
n
palladium n-alkyl product 3 and iso-alkyl isomer 3 are consid-
erably stabilized (22.8 kcalmolꢀ1 and 23.1 kcalmolꢀ1) by the
dissociation of the weakly coordinating HOTFA molecule and
Scheme 1. Proposed mechanism for the aerobic partial oxidation of pro-
pane. VV: NaVO3. Vred: vanadium-oxo compound with an oxidation state
lower than +V.
n
the formation of a b agostic interaction in the intermediates 4
or iso4.
Very recently, we discovered that the reaction with propane
can be run with molecular oxygen as terminal oxidant in the
presence of a NaVO3 co-catalyst and bromide.[14] Our results
therefore constitute an example of aerobic generation of palla-
dium in the oxidation state of +IV.[15,16] DFT calculations[17] sug-
gested that a mixture of neutral and charged species are pres-
ent under the reaction conditions prior to the formation of the
catalytically competent species 1.
Herein, we report detailed mechanistic investigations, includ-
ing DFT calculations[18] on the CH functionalization of propane
by palladium bis(NHC) catalysts. We will rationalize why the re-
action proceeds with iso-selectivity and report the selectivity
and rate determining step of the reaction. Calculated reaction
barriers will be shown to correlate with the catalytic activity of
a series of bis(NHC) complexes.[11] We will comment on the ki-
netics of the reaction and on the stoichiometric reactivity of
potential intermediates of the catalytic cycle.
Oxidation and product formation
Bromine can replace the agostic interaction in 4 and iso4 by
coordinating to the palladium atom in compounds 5 and iso5.
n
n
The oxidative addition of bromine to produce n6 or iso6
(Scheme 3, Figure 2) then proceeds via the transition states
ts[n5>n6] (14.1 kcalmolꢀ1) or ts[iso5>iso6] (11.1 kcalmolꢀ1).
The kinetic preference for oxidation of the iso-intermediate
iso5 is also reflected by the greater thermodynamic stability of
the palladium(IV) product iso6 (ꢀ31.5 kcalmolꢀ1) versus n6
(ꢀ26.0 kcalmolꢀ1). Compounds iso6 and 6 can then easily re-
n
ductively eliminate the corresponding n- or iso- trifluoroacetic
acid ester.
Results and Discussion
CH Activation
The CH activation (Scheme 2) is predicted to take place after
the endergonic formation of the cationic, propane-coordinat-
Scheme 3. Oxidation and reductive elimination. Values are given in kcal
molꢀ1
.
Figure 2 illustrates that the bond length of the forming Pdꢀ
Br bond in the iso transition state ts[iso5>iso6] is longer by
0.06 ꢁ compared to the n transition state ts[n5>n6]. The pref-
erence for the iso-transition state is mainly due to electronic
reasons. The oxidation of the palladium(II) complex to the cor-
responding palladium(IV) compound entails an increasing posi-
tive partial charge on the palladium atom (natural populations
analysis (NPA)[19] charges, values given in atomic units: iso5: +
Scheme 2. CH Activation at the iso- and n-position of propane. Values are
given in kcalmolꢀ1. Vv and Vred are defined as in Scheme 1.
n
n
0.43 a.u.; iso6: +0.59 a.u.; 5: +0.43 a.u.; 6 +0.66 a.u.). As the
Chem. Eur. J. 2014, 20, 14872 – 14879
www.chemeurj.org
14873 ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
merization process efficiently. This is in line with our experi-
mental observation of the predominant formation of the iso-
product (see below in Figure 7 and Supporting Information
Figures S2–S4). Consequently, reaction mechanisms involving
the reductive elimination from palladium in the oxidation
state +IV are predicted to preferentially yield the iso-function-
alized product.
Discussion of the DFT calculated mechanism
Figure 2. Oxidative addition transition states ts[iso5>iso6] and ts[n5>iso6].
Scheme 4 shows the DG (and DH) values for the most favora-
ble calculated mechanism. The catalytic cycle starts with the
preactivation of 7 by oxidative substitution of the strongly co-
ordinating bromido for weakly coordinating trifluoroacetato li-
gands (1, DG=ꢀ52.1 kcalmolꢀ1). The CH activation step takes
place at one of the CH3 groups of propane and produces the
n-alkyl palladium species n4. This alkyl palladium species iso-
merizes to the corresponding iso-propyl compound iso4. The
oxidation of compound iso4 to palladium(IV) is mediated by
bromine and followed by the reductive elimination of the iso-
propyl ester. The oxidation reaction proceeds with an overall
reaction barrier of 34.2 kcalmolꢀ1, whereas the CH activation
step is predicted to have a reaction barrier of 38.1 kcalmolꢀ1.
Therefore, the regioselectivity of the overall reaction (where
oxidation is the selectivity-determining step) is not established
within the rate-determining step (CH activation). The calculat-
ed rate-determining transition state predicts a kinetic isotope
effect (KIE) between perprotio and perdeuterio propane of 4.1
at 608C. A comparison of the DG and DH values reveals that
the reaction proceeds with a negative entropy (DS <0). Never-
theless, considering the uncertainties related to the calculation
of DS values of processes associated with ligand dissociation
and charge separation, we do not want to put too much
weight on the calculated absolute values.[21] Electrophilic
mechanisms without the intermediate formation of carbon–
Bond lengths are given in [ꢁ], NPA charges in a.u.
iso-propyl substituent is a stronger electron donor than the n-
propyl substituent, the corresponding transition state ts[iso5>
iso6] shows an enhanced stabilizing charge transfer from the
bonding carbon atom to the transition metal (NPA charges:
ts[iso5>iso6]: ꢀ0.30 a.u.; ts[n5>n6]: ꢀ0.55 a.u.). Most interest-
ingly, this umpolung process[20] facilitates the nucleophilic
attack by trifluoroacetate leading finally to the liberation of
iso-propyl trifluoroacetate.
Isomerization
n
Compound 4 can interconvert to iso4 via an isomerization/mi-
gratory insertion reaction sequence (Supporting Information,
Figure S1). b-Hydride elimination in intermediate n4 leads to
the formation of a propene-coordinating p-complex 4_p. Rota-
tion around the propene double bond (ts[n4>iso4]) and migra-
tory insertion of the hydride atom produces compound iso4.
This reaction sequence proceeds with a relatively low reaction
¼
barrier (DG <5 kcalmolꢀ1). We conclude that the different n-
and iso-alkyl compounds can potentially exist in equilibrium
with each other. Even chelating bis(NHC) ligands with bulky
tert-butyl groups seem to be incapable of preventing this iso-
Scheme 4. Most favorable calculated mechanism. DG (DH) values are given in kcalmolꢀ1
Chem. Eur. J. 2014, 20, 14872 – 14879 www.chemeurj.org
.
14874
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
transition metal bonds have been investigated as well, but
were predicted to have prohibitively high reaction barriers
clude that the oxidant is not involved in the rate-determining
step. The reaction furthermore exhibits a dependence on a 1st
order reaction rate relative to the catalyst’s and substrate’s
concentrations, as predicted by the calculations.[14a] In conclu-
sion, all the obtained kinetic data and rate laws agree with the
proposed mechanism.
¼
(DG >52 kcalmolꢀ1, Figure S2 in the Supporting Information).
The combined results are in line with the literature, as the n-
selectivity, which is often observed in case of transition-metal-
mediated CH activation chemistry, has been reported to be
very small or even reversed in the case of (especially cationic)
palladium(II) catalysis.[22] Very detailed DFT calculations report-
ed by Harvey rationalized the preferential formation of the iso
branched product for chelating palladium diimine complexes
with electronic reasons.[23] Notably, Harvey also described an
enhanced selectivity for the iso-product in case of cationic pal-
ladium complexes.
Formation of CO2
We detected the formation of carbon dioxide during the
course of the reaction, which can be attributed to the decar-
boxylation of trifluoroacetic acid. The amount of CO2 formed
accounts roughly for the observed low mass balance in K2S2O8.
We also determined a loss of about 1–2 mL of solvent in the
kinetic experiment shown in Figure 3. Hence, our results indi-
cate that radical oxidation of the solvent by K2S2O8 to CO2
occurs. This pathway, which has been proposed to lead to CH
activation by the generation of CF3 radicals[26] is definitely not
involved in the case of the oxidants Selectfluor or NaVO3/O2,
where we did not detect the formation of CO2.[14a] Also, this
pathway does not contribute considerably to the formation of
the trifluoroacetate ester under our reaction conditions with
K2S2O8, as no effect of adding radical transfer/trapping re-
agents to the reaction was observed.[14]
In order to experimentally evaluate the plausibility of our
calculated rate-determining step, we determined the Arrhenius
parameters and the isotope effect of the reaction.[14a] The Ar-
rhenius energy of activation was determined to be (29.0ꢁ
1.6) kcalmolꢀ1, which fits nicely to the calculated DH value of
30.0 kcalmolꢀ1. Also the isotope effect of 3.4ꢁ0.8, which was
obtained for the model substrate cyclohexane, is in good
agreement with calculated values (cyclohexane=3.9, pro-
pane=4.1). Both experimental values confirm that the CH-acti-
vation step is the rate determining state of the reaction. We
also observed that the addition of water and even of trifluoro-
acetic anhydride led to an inhibition of the reaction. This can
also be rationalized by our calculations,[12] as the rate-determin-
ing transition state is cationic in nature.[24] The DFT-calculated
mechanism has a number of further implications, which could
be verified by experimental investigations. The reaction is ex-
pected to proceed without an induction period and with
a 1st order rate law regarding the catalyst’s concentration. A
zero order rate law is expected for a soluble[25] oxidant, as it is
not directly involved in the rate-determining step (CH activa-
tion). Furthermore, the activity of catalysts with different bis-
(NHC) ligands should be correlated to the calculated barriers
for the rate determining state.
Alkylbromides
The DFT-calculated mechanism involves the reductive elimina-
tion of iso-propyl trifluoroacetate from a palladium(IV) inter-
mediate with bromido ligands. Our calculations for methane
had predicted that the reductive elimination by reaction with
trifluoroacetate is favored over reductive elimination of the
corresponding bromoalkyl compound.[12] Nevertheless, we
wanted to verify if alkylbromides are indeed not viable inter-
mediates of the catalytic cycle. Therefore, 2-bromopropane
was subjected to the reaction conditions (Figure 4). However,
as it did not react, there is no reason to believe that it is an in-
termediate of the catalytic cycle.
Rate laws
The kinetics of the reaction was studied. We did not observe
a significant induction period or signs of decomposition of the
palladium catalyst after a reaction time of 23 h. A plot of the
reaction kinetics (Figure 3) confirms that the reaction is not de-
pendent on the concentration of oxidant (approximately zero
order) for the first 12 h (ꢂ50% yield based on oxidant) at an
oxidant loading of 400 equiv to the catalyst. We therefore con-
Figure 4. 2-Bromopropane is not a viable intermediate of the catalytic cycle.
Reaction conditions: 1.3 mmol 7, 2.7 mmol K2S2O8, 50 mL 2-bromopropane,
0.8 mL HOTFA, 608C, 17 h.
Next, we turned our attention towards the effect of adding
bromide salts to the reaction. As already reported,[14] the addi-
tion of two or five equivalents of KBr to the reaction mixture
led to an almost quantitative inhibition of the reaction with
very little (supposedly radical) background reactivity (4.2 mmol
7, 8.4 mmol or 21 mmol KBr, 210 mmol K2S2O8, 1 bar propane,
2.5 mL HOTFA, 608C, 17 h: TON=4 and TON=2, compared to
TON=18 without addition of KBr). Surprisingly, we also ob-
served the formation of iso-propyl bromide and iso-propyl tri-
Figure 3. Kinetic profile. Reaction conditions: 113 mmol 7, 400 equiv K2S2O8,
67.5 mL HOTFA, 608C, 6 bar propane.
Chem. Eur. J. 2014, 20, 14872 – 14879
www.chemeurj.org
14875
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 5. Autocatalysis at high loadings of KBr. 210 mmol K2S2O8, 2.5 mL
HOTFA, 1 bar propane, 608C, 17 h.
Figure 6. Formation of L2PdIV(X)4 after addition of oxidant as evidenced by
1H NMR spectroscopy. Upper spectrum: Signals of 7 after addition of K2S2O8
in HOTFA. Lower spectrum: Signals of 7 after addition of Br2 in HOTFA with
assignments of signals for 7.
fluoroacetate, if we further increased the amount of KBr to
eight equivalents (32 mmol, TON=11). Further investigations
revealed that the reactivity is not associated with the palladi-
um catalyst (Figure 5).
1
ligand) as described by Kraft.[13d] A H NMR spectrum revealed
The reaction proceeds with extraordinary chemo- and regio-
selectivity. We did not detect the formation of n-alkyl, difunc-
tionalized, or any other products, but the formation of moder-
ate amounts of 2-bromopropane at KBr loadings of 42 mmol or
higher. This points towards a different reaction mechanism
than for the palladium catalysts. Remarkably, the maximum
yield was obtained at a KBr/K2S2O8 ratio of 0.5:1. The virtually
quantitative iso-selectivity could be explained by a mechanism
involving electrophilic, positively charged bromo species. Relat-
ed I+ reactivity was proposed for the oxidation of methane in
oleum[27] and was also discussed for LiBr in the presence of pe-
riodates.[3f,28] Very recent work demonstrated that the presence
of chloride salts is decisive for the very efficient and selective
oxidation of methane to methyl trifluoroacetate.[29] We con-
clude that this area of hydrocarbon functionalization looks
very promising, but does not appear to be the preferential
pathway in the presence of low concentrations of 7 or for pal-
ladium acetate.
the existence of at least two different bis(NHC) complexes in
solution (Figure 6). Apart from the starting material, it showed
signals at 4.17 (CH3), 6.98 (CH2), 7.15 (CH2), 7.56 (CH2) and
a broad signal at 8.01 ppm (CH). When adding bromine to a so-
lution of 7 in trifluoroacetic acid, we observed the formation of
a very similar signal set, which can be assigned to a palladium
species in the oxidation state +IV.[13b,d] Accordingly, the addi-
tion of not only bromine, but also of K2S2O8, to 7 in trifluoro-
acetic acid, leads to the formation of a mixture of palladium(II)
and palladium(IV) compounds. This mixture constitutes the
resting state of the catalyst.
Finally, the catalytic activity of the palladium(IV) complex
L2PdIV(X)4 was investigated. However, no significant reactivity
was obtained even after the addition of silver triflate. In agree-
ment with the DFT calculations, CH activation by palladium(IV)
species does not appear to be part of the operating reaction
mechanism.
Stoichiometric reactivity
Structure–activity relationship
The stoichiometric reactivity of the palladium(II) complexes 7
and 1 with propane and cyclohexane was evaluated. The DFT
calculations predict 7 and 1 to be catalytically inactive, as the
CH functionalization is proposed to be mediated by a cationic
palladium(II) complex. The resulting intermediate needs to be
oxidized subsequently by an oxidant. Indeed, we did not ob-
serve the formation of alkyl trifluoroacetates or alkyl halides,
and the complexes appeared to be perfectly stable in refluxing
trifluoroacetic acid in the presence of propane. Also, the in situ
exchange of the bromido ligands by two equivalents of AgOTf
did not result in the formation of trifluoroacetoxylated prod-
ucts. We conclude that, contrarily to the reactivity of Pd-
(OAc)2,[8c] activation by palladium(II), followed by reductive
After having established the nature of the rate determining
step, we evaluated whether or not we could rationalize the
catalytic activity of palladium catalysts with different bis(NHC)
ligands. Indeed, the calculations turned out to predict the ex-
perimental results fairly well. Figure 7 shows the correlation be-
tween the calculated overall Gibbs free energy of activation
and the catalytic turnover (TON) after 12 h for an extended[11]
series of palladium bis(NHC) catalysts. Less-electron-rich benzi-
midazolylidene catalysts are predicted to show considerably
lower activity than their imidazolylidene counterparts, which is
in agreement with our experimental data. The same applies for
the ortho-phenylene bridged catalysts with bromo substituents
at the phenyl ring compared with substitution with hydrogen
substituents. Very impressively, even the small steric effects of
the tert-butyl, methyl, or benzyl ligand wingtips are rational-
ized by the calculations. Therefore, the calculations could serve
as a predictive tool for the further development of more active
and selective catalysts, as long as the rate-determining state
does not change.
elimination to palladium(0), is not
mechanism.
a
viable reaction
In order to prove the formation of palladium(IV) species in
the reaction mixture, we added K2S2O8 to a solution of 7 in
HOTFA (14 h, 608C). We thereby observed a reddish precipi-
tate, which we attribute to the formation of unstable and in-
soluble palladium(IV) complexes L2PdIV(X)4 (L2 =bis(NHC)
Chem. Eur. J. 2014, 20, 14872 – 14879
www.chemeurj.org
14876
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
¼
Figure 7. Correlation between the observed catalytic activity (TON) and the reaction barrier (DG ) as predicted by DFT (i.e., 1 and ts[n2>n3]). Reaction condi-
tions: 8.4 mmol catalyst, 0.84 mmolK2S2O8, 5.0 mL HOTFA, 608C, 12 h. Values are given in kcalmolꢀ1
.
All palladium bis(NHC) complexes were synthesized as reported in
the literature.[10c,11,31] All chemicals were obtained by commercial
suppliers and used without further purification. KBr with a purity
of 99.999% has been used.
Conclusion
The mechanism of the bis(NHC) palladium complex catalyzed
functionalization of propane was studied in detail. We report
and discuss DFT calculations, kinetic data, stoichiometric reac-
tivity, rate laws, solvent effects, ligand–catalytic activity rela-
tionship, and the activation energy. The experimental results
are in agreement with the predictions by DFT. There is
a strong indication that the CH-activation step takes place at
a palladium(II) complex and is followed by an oxidation by
bromine to the oxidation state +IV. The combined results of
DFT calculations and experiments are indicative that the CH-ac-
tivation step controls the reaction rate, whereas the iso-selec-
tivity is being determined within the oxidation step to palla-
dium(IV). A different mechanism with remarkably high chemo-
selectivity was shown to be operative at high loadings of bro-
mine salts. We believe that our DFT calculations will guide fur-
ther improvements regarding the ligand design and will
support the development of novel aerobic cooxidation
procedures.
The catalyses described in Figure 7 were conducted as previously
reported[11] and were run in 10 mL crimp-top vials with PTFE-faced
butyl rubber septa. Palladium (8.4 mmol) catalyst and K2S2O8
(0.84 mmol) were suspended in HOTFA (5 mL). In cases where the
catalyst amount was less than 5 mg, stock solutions of the catalyst
in HOTFA were freshly prepared. Subsequently, propane was bub-
bled through the suspension for 60 seconds. The vials were sealed
immediately and it was stirred vigorously at 608C for 12 h. The re-
action mixture was analyzed by gas chromatography with an Agi-
lent 6850 Series II Networked GC equipped with a flame ionisation
detector (FID) and a Macherey–Nagel Optima-210 0.25 mm column
(30 mꢂ0.25 mm). Quantification of the yield of the reaction prod-
ucts was done by adding C6F5Cl (25 mL) as standard to the reaction
mixture (1.00 mL) and injection of this mixture into the GC. Report-
ed TONs are based on iso-propyl trifluoroacetate only. The identity
of the reaction products was confirmed by comparison with the re-
1
tention time of defined samples and H NMR spectra of the reac-
tion mixtures in an NMR tube, which contained a capillary filled
with [D6]DMSO to provide a lock signal. All GC measurements were
repeated four times and catalyses twice. Given values are averaged
over all measurements. For the examination of the reactivity of iso-
propyl bromide, 1.3 mmol 7, 2.7 mmol K2S2O8 and 50 mL 2-bromo-
propane were added to 0.8 mL HOTFA in the 10 mL crimp top vial,
which was sealed subsequently. The reaction mixture was analyzed
by GC-FID after stirring for 17 h at 608C.
Experimental Section
DFT Calculations
The calculation have been performed with Gaussian03 at the SP//
B3LYP/6–31G(d) level of theory using a Hay–Wadt effective core
potential (SP: B3LYP-D3[SCRF(CPCM)]/6–311+ +G(d,p)) and ac-
cording to our previous reports.[12,17,30] Further details and the full
Gaussian citation are given in the Supporting Information.
For the determination of the kinetics (Figure 3), 113 mmol of 7 and
45.2 mmolK2S2O8 were added to 67.5 mL of HOTFA in a 160 mL
Hastelloy C-2000 reactor by Parr. The reactor was purged three
times with 7 bar of propane and then filled with 7 bar of propane.
It was stirred at 500 rpm and 608C. Samples were taken in situ
with the help of a sampling valve and analyzed by GC-FID at the
indicated time points.
Experimental details
Caution: Many of the reagents and reaction conditions are poten-
tially hazardous.
The analysis of the gas headspace was performed with an Agilent
6850 Series II Networked GC, equipped with a thermal conductivity
detector (TCD) and a J&W Scientific GS-GasPro column (60 mꢂ
Proper safety precautions should be implemented!
Chem. Eur. J. 2014, 20, 14872 – 14879
www.chemeurj.org
14877
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[8] a) A. Sen, Top. Organomet. Chem. 1999, 3, 81–95; b) C. Jia, T. Kitamura,
Y. Fujiwara, Acc. Chem. Res. 2001, 34, 633–639; c) N. F. Goldshleger, A. P.
Moravskii, Russ. Chem. Rev. 1994, 63, 125–138; d) A. N. Vedernikov, A. I.
Kuramshin, B. N. Solomonov, J. Chem. Soc. Chem. Commun. 1994, 121–
122.
0.32 mm). Gas samples were taken from the reactor at room tem-
perature with a custom build 1.5 mL high pressure gas mouse
after the reaction. The gas sample was then expanded into an
open-ended 1 mL sample loop at the GC. Identification and quan-
tification of the detected gases was accomplished by comparing
retention times and signal areas with measurements of standard
gas mixtures with known gas compositions.
[9] M. Golombok, T. Nijbacker, M. Raimondi, Ind. Eng. Chem. Res. 2004, 43,
6001–6005.
[10] a) M. Muehlhofer, T. Strassner, W. A. Herrmann, Angew. Chem. 2002, 114,
1817–1819; Angew. Chem. Int. Ed. 2002, 41, 1745–1747; b) D. Meyer,
M. A. Taige, A. Zeller, K. Hohlfeld, S. Ahrens, T. Strassner, Organometallics
2009, 28, 2142–2149; c) S. Ahrens, A. Zeller, M. Taige, T. Strassner, Orga-
nometallics 2006, 25, 5409–5415; d) T. Strassner, M. Muehlhofer, A.
Zeller, E. Herdtweck, W. A. Herrmann, J. Organomet. Chem. 2004, 689,
1418–1424.
For the stoichiometric oxidation of 7, the catalyst (31 mmol) was
dissolved in a screw cap NMR tube in trifluoroacetic acid (0.36 mL).
Trifluoroacetic acid (0.04 mL) was added in order to guarantee
water-free conditions. The NMR tube contained a capillary filled
with [D6]DMSO in order to provide a lock signal. The NMR tube
was sealed quickly after suspending K2S2O8 (31 mmol) and heated
overnight at 608C, where it was kept under constant vibration. The
spectra were recorded on a Bruker AC 300P spectrometer.
The experiment with L2PdIV(Cl)4 was conducted accordingly, but
AgOTf was added instead of K2S2O8 and propane was bubbled
through the reaction mixture for 30 seconds prior to sealing the
screw cap NMR tube. The reactivity of 7 and 1 with propane was
studied likewise. For cyclohexane, 28 mL (256 mmol) were added to
the reaction mixture instead of saturating it with propane.
[11] D. Munz, A. Poethig, A. Tronnier, T. Strassner, Dalton Trans. 2013, 42,
7297–7304.
[12] D. Munz, D. Meyer, T. Strassner, Organometallics 2013, 32, 3469–3480.
[13] a) D. Meyer, Dissertation Thesis, TU Dresden (Dresden, Germany), 2011;
b) D. Meyer, S. Ahrens, T. Strassner, Organometallics 2010, 29, 3392–
3396; c) A. S. McCall, S. Kraft, Organometallics 2012, 31, 3527–3538;
d) A. S. McCall, H. Wang, J. M. Desper, S. Kraft, J. Am. Chem. Soc. 2011,
133, 1832–1848; e) S. S. Subramanium, L. M. Slaughter, Dalton Trans.
2009, 6930–6933; f) S. N. Sluijter, S. Warsink, M. Lutz, C. J. Elsevier,
Dalton Trans. 2013, 42, 7365–7372; g) M. Finger, Dissertation Thesis,
Universität Heidelberg (Germany), 2008.
[14] a) D. Munz, T. Strassner, Angew. Chem. 2014, 126, 2518–2521; Angew.
Chem. Int. Ed. 2014, 53, 2485–2488;
Acknowledgements
[15] a) K. M. Engle, T. S. Mei, X. Wang, J. Q. Yu, Angew. Chem. 2011, 123,
1514–1528; Angew. Chem. Int. Ed. 2011, 50, 1478–1491; b) D. C.
Powers, T. Ritter, Acc. Chem. Res. 2012, 45, 840–850; c) L. Ackermann,
Chem. Rev. 2011, 111, 1315–1345; d) C. S. Yeung, V. M. Dong, Chem. Rev.
2011, 111, 1215–1292; e) P. Sehnal, R. J. Taylor, I. J. Fairlamb, Chem. Rev.
2010, 110, 824–889.
[16] a) A. N. Vedernikov, Acc. Chem. Res. 2012, 45, 803–813; b) J. R. Khusnut-
dinova, F. R. Qu, Y. Zhang, N. P. Rath, L. M. Mirica, Organometallics 2012,
31, 4627–4630; c) L. Boisvert, K. I. Goldberg, Acc. Chem. Res. 2012, 45,
899–910; d) A. N. Campbell, S. S. Stahl, Acc. Chem. Res. 2012, 45, 851–
863.
[17] D. Munz, T. Strassner, Top. Catal. 2014, DOI: 10.1007/s11244-014-0305-5.
[18] a) D. Balcells, E. Clot, O. Eisenstein, Chem. Rev. 2010, 110, 749–823; b) Y.
Boutadla, D. L. Davies, S. A. Macgregor, A. I. Poblador-Bahamonde,
Dalton Trans. 2009, 5820–5831; c) K. J. Bonney, F. Schoenebeck, Chem.
Soc. Rev. 2014, DOI: 10.1039/c1034s00061g.
We thank the Center for Information Services and High Per-
formance Computing (ZIH) for computation time and the
Deutsche Forschungsgesellschaft (DFG) (STR 526/7-1 and STR
526/7-2) for financial support. D.M. thanks the Studienstiftung
des Deutschen Volkes for support. Donations of trifluoroacetic
acid by Solvay Fluor are gratefully acknowledged.
Keywords: CꢀH activation · DFT · alkanes · NHC · palladium
[1] B. G. Hashiguchi, M. M. Konnick, S. M. Bischof, S. J. Gustafson, D. Devara-
jan, N. Gunsalus, D. H. Ess, R. A. Periana, Science 2014, 343, 1232–1237.
[2] BP Energy Outlook 2013, London, 2001.
[19] A. E. Reed, L. A. Curtiss, F. Weinhold, Chem. Rev. 1988, 88, 899–926.
[20] J. A. Labinger, J. E. Bercaw, Top. Organomet. Chem. 2011, 35, 29–59.
[21] P. Deglmann, S. Schenk, J. Comput. Chem. 2012, 33, 1304–1320.
[22] a) L. H. Shultz, D. J. Tempel, M. Brookhart, J. Am. Chem. Soc. 2001, 123,
11539–11555; b) A. Michalak, T. Ziegler, Organometallics 2000, 19,
1850–1858; c) A. Michalak, T. Ziegler, Organometallics 1999, 18, 3998–
4004; d) D. L. Reger, D. G. Garza, Organometallics 1993, 12, 554–558;
e) D. L. Reger, Y. Ding, D. G. Garza, L. Lebioda, J. Organomet. Chem.
1993, 452, 263–270; f) D. L. Reger, D. G. Garza, L. Lebioda, Organometal-
lics 1992, 11, 4285–4292; g) D. L. Reger, D. G. Garza, L. Lebioda, Organo-
metallics 1991, 10, 902–906.
[23] J. N. Harvey, Organometallics 2001, 20, 4887–4895.
[24] M. Ahlquist, R. A. Periana, W. A. Goddard, Chem. Commun. 2009, 2373–
2375.
[25] J. B. Gary, A. K. Cook, M. S. Sanford, Acs Catal. 2013, 3, 700–703.
[26] D. M. Camaioni, J. T. Bays, W. J. Shaw, J. C. Linehan, J. C. Birnbaum, J.
Org. Chem. 2001, 66, 789–795.
[27] a) R. A. Periana, O. Mirinov, D. J. Taube, S. Gamble, Chem. Commun.
2002, 2376–2377; b) X. Gang, Y. M. Zhu, H. Birch, H. A. Hjuler, N. J. Bjer-
rum, Appl. Catal. A 2004, 261, 91–98.
[3] a) J. A. Labinger, J. Mol. Catal. A 2004, 220, 27–35; b) R. H. Crabtree, J.
Organomet. Chem. 2004, 689, 4083–4091; c) M. Sun, J. Z. Zhang, P.
Putaj, V. Caps, F. Lefebvre, J. Pelletier, J. M. Basset, Chem. Rev. 2014, 114,
981–1019; d) V. N. Cavaliere, D. J. Mindiola, Chem. Sci. 2012, 3, 3356–
3365; e) G. A. Olah, G. K. S. Prakash, Electrophilic Reactions of Alkanes.
The Chemistry of Alkanes and Cycloalkanes, Wiley, Hoboken, 2006; f) I.
Akhrem, A. Orlinkov, Chem. Rev. 2007, 107, 2037–2079; g) E. G. Chepai-
kin, Russ. Chem. Rev. 2011, 80, 363–396; h) H. Schwarz, Angew. Chem.
2011, 123, 10276–10297; Angew. Chem. Int. Ed. 2011, 50, 10096–10115;
i) J. R. Webb, T. Bolano, T. B. Gunnoe, ChemSusChem 2011, 4, 37–49.
[4] A. E. Shilov, G. B. Shulpin, Activation and catalytic reactions of saturated
hydrocarbons in the presence of metal complexes, Kluwer Academic, Dor-
drecht, 2002.
[5] B. G. Hashiguchi, S. M. Bischof, M. M. Konnick, R. A. Periana, Acc. Chem.
Res. 2012, 45, 885–898.
[6] a) T. Strassner, S. Ahrens, M. Muehlhofer, D. Munz, A. Zeller, Eur. J. Inorg.
Chem. 2013, 3659–3663; b) M. N. Vargaftik, I. P. Stolarov, I. I. Moiseev, J.
Chem. Soc. Chem. Commun. 1990, 1049–1050; c) W. Chen, J. A. Kocal,
T. A. Brandvold, M. L. Bricker, S. R. Bare, R. W. Broach, N. Greenlay, K.
Popp, J. T. Walenga, S. S. Yang, J. J. Low, Catal. Today 2009, 140, 157–
161.
[7] a) M. V. Kirillova, A. M. Kirillov, M. L. Kuznetsov, J. A. L. Silva, J. J. R. F. da
Silva, A. J. L. Pombeiro, Chem. Commun. 2009, 2353–2355; b) M. V. Kiril-
lova, J. A. L. da Silva, J. J. R. Frafflsto da Silva, A. J. L. Pombeiro, Appl.
Catal. A 2007, 332, 159–165; c) M. V. Kirillova, M. L. Kuznetsov, P. M.
Reis, J. A. da Silva, J. J. da ilva, A. J. Pombeiro, J. Am. Chem. Soc. 2007,
129, 10531–10545; d) G. B. Shul’pin, Mini-Rev. Org. Chem. 2009, 6, 95–
104; e) A. A. Fokin, P. R. Schreiner, Chem. Rev. 2002, 102, 1551–1594.
´
´
[28] a) M. Jarosinska, K. Lubkowski, J. G. Sosnicki, B. Michalkiewicz, Catal.
Lett. 2008, 126, 407–412; b) T. M. Shaikh, A. Sudalai, Tetrahedron Lett.
2005, 46, 5589–5592; c) G. A. Olah, P. Schillin, J. Am. Chem. Soc. 1973,
95, 7680–7686.
[29] G. C. Fortman, N. Boaz, D. Munz, M. M. Konnick, R. A. Periana, J. Groves,
T. B. Gunnoe, J. Am. Chem. Soc. 2014, 136, 8393–8401.
[30] a) Gaussian 03, Rev. E.01, M. J. Frisch et al., Gaussian, Inc., Pittsburgh,
PA, 2003; b) A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652; c) C. Lee,
Chem. Eur. J. 2014, 20, 14872 – 14879
www.chemeurj.org
14878
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
W. Yang, R. G. Parr, Phys. Rev. B: Condens. Matter 1988, 37, 785–789;
d) P. J. Stephens, F. J. Devlin, C. F. Chabalowski, M. J. Frisch, J. Phys.
Chem. 1994, 98, 11623–11627; e) S. H. Vosko, L. Wilk, M. Nusair, Can. J.
Phys. 1980, 58, 1200–1211; f) W. J. Hehre, R. Ditchfield, J. A. Pople, J.
Chem. Phys. 1980, 56, 2257–2261; g) W. R. Wadt, P. J. Hay, J. Chem. Phys.
1985, 82, 284–298; h) R. C. Binning, Jr., L. A. Curtiss, J. Comput. Chem.
1990, 11, 1206–1216; i) S. Grimme, J. Antony, T. Schwabe, C. Muck-Lich-
tenfeld, Org. Biomol. Chem. 2007, 5, 741–758; j) S. Grimme, J. Comput.
Chem. 2006, 27, 1787–1799; k) A. D. McLean, G. S. Chandler, J. Chem.
Phys. 1980, 72, 5639–5648; l) R. Krishnan, J. S. Binkley, R. Seeger, J. A.
Pople, J. Chem. Phys. 1980, 72, 650–654; m) L. A. Curtiss, M. P. McGrath,
J. P. Blaudeau, N. E. Davis, R. C. Binning, L. Radom, J. Chem. Phys. 1995,
103, 6104–6113; n) V. Barone, M. Cossi, J. Phys. Chem. A 1998, 102,
1995–2001; o) M. Cossi, N. Rega, G. Scalmani, V. Barone, J. Comput.
Chem. 2003, 24, 669–681; p) S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J.
Chem. Phys. 2010, 132, 154104; q) S. Grimme, S. Ehrlich, L. Goerigk, J.
Comput. Chem. 2011, 32, 1456–1465.
[31] a) W. A. Herrmann, J. Schwarz, M. G. Gardiner, Organometallics 1999, 18,
4082–4089; b) E. Herdtweck, M. Muehlhofer, T. Strassner, Acta Crystal-
logr. Sect. E 2003, 59, m970–m971; c) W. Herrmann, T. Scherg, S.
Schneider, G. Frey, J. Schwarz, E. Herdtweck, Synlett 2006, 2894–2907.
Received: June 11, 2014
Published online on September 24, 2014
Chem. Eur. J. 2014, 20, 14872 – 14879
www.chemeurj.org
14879
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim